Design Challenge (Was Sunny Island Experience?) [RE-wrenches]

Jeff Clearwater clrwater at earthlink.net
Mon Feb 6 12:29:46 PST 2006


<x-flowed>
Hi William,

So nice of you to delve into thinking about this . . . .

Yes I've been thinking along these lines too.   I need to run right 
now -  and nee to put more thought into it and drawit out more but is 
not the issue the fact that you lose the inverter's genset charge 
controls and auto start if you put it on the output side of the 
inverters?  Or does one just loop the outputs and inputs - but then 
the charging software can't sense just gen inputs - so like you say 
you'ld have to control in other ways.  Am I on base here? 
Interesting . . .

The European diagram of the Sunny Island's AC coupling differs from 
the U.S. version in just this way (or maybe it's just a graphic 
convenience) but they show the genset on the AC bus with everything 
else.  In the U.S. version the Genset clearly feeds the inputs with 
everything else on the outputs.  Sam, John?  help us out here.

If the Sunny Island (or future off-grid Outbacks) had that ability - 
to have a single AC input/output for charging and loads - then we 
really start to benefit from this AC grid stuff in terms of generator 
placement.

Thanks again for your thoughts - any more is more than welcome!

Jeff C.
Village Power Design



>
>A thought here... is the generator capable of self-synchronizing its 
>output? If it can sync to the inverters' output, here's a 
>configuration that might cut the wire cost and complexity...
>
>Solar array on grid-tie inverters located at corner A of triangle
>(A connected to B with (3) 3/0 aluminum sets or single 500 MCM set)
>Houses/loads, and battery Inverters on corner B of triangle
>(B connected to C with (3) 3/0 aluminum sets or single 500 MCM set)
>Generator connected on corner C of triangle
>
>All power components are coupled through a common 240Vac AC bus...
>
>-Use grid tie inverters for all of the PV output. Use a battery 
>voltage monitor to shut off SMA inverters (staged relays) on high 
>batt voltage.
>-Operate the generator by two signal inputs: (1) operate it when 
>battery voltage drops below a determined threshold (2) operate it 
>when load on the battery inverters exceeds a determined threshold 
>(say, 40A for 15 seconds measured by a load monitoring sensor).
>
>With this method, any grid tie inverter output (PV production) would 
>automatically displace demand on the battery inverter output, 
>increasing the capacity to handle peak loads during daytime hours. 
>Power flows run the shortest possible distance and at reasonably 
>convenient voltage.
>
>Average load voltage drop for PV to load is under 1%, leaving room 
>for 2-3x expansion
>Voltage drop Generator to load = 2.5% @120A, or 1.8% @more typical 20kW.
>
>The loads past the generator spot should have acceptable voltage 
>drop in all cases so long as the feeder from generator to them is 
>sized ~1% drop
>
>Have customer try to run loads during PV output hours and they 
>should be able nearly avoid use of the generator for peak load 
>demands.
>
>Maybe I'll think about this some more...
>
>/wk
>
>Jeff Clearwater wrote:
>>Hi Good Colleagues,
>>
>>Thanks for the feedback Jeff and Travis - good to hear AC coupling 
>>is out there and being used successfully in the US!
>>
>>OKAY DESIGN MINDS - Here's my design challenge - input most 
>>appreciated!  Consolation prizes available after the contest (just 
>>kidding).
>>
>>I'm designing a multi-residence off-grid system with 12 KW PV, 30 
>>KW (120 A) Diesel Genset, 12-15 KWs of inverter loads with about 24 
>>KW peak, a 100 KWH battery pack and the following long distances 
>>between components:
>>
>>Picture an equilateral triangle with roughly 600 foot sides. Point 
>>1 is Loads, Point 2 is  Generator, Point 3 is PV array site.  This 
>>is pretty close to the actual situation on the ground (as 
>>unfortunate as that might be). Close enough for easy discussions. 
>>Power Shed has yet to be located:  Generator site is practically 
>>set in stone. There are also some loads past the Genset side of the 
>>triangle.
>>
>>In the future we will also have wind and hydro coming in from about 
>>2000' away from the PV side of the triangle.  Generator is in and 
>>can't be moved, neither can the house complex, and the PV site is 
>>the only one non-shaded so it's a given too.
>>
>>PROPOSED SYSTEM #1:  4 Sunny Islands (2 sets of 2 stacked split 
>>phase) in power shed located near the generator shed.  This allows 
>>us to feed the long distance array to the system at high voltage - 
>>either high voltage DC or 240 VAC depending on where we located the 
>>SB.  It also allows us to not have a ditch and conduit along the 
>>PV<>Power Shed side of the triangle as the PV can feed into the AC 
>>line at the house.  It also allows us to to feed Windy Boys and 
>>Hydro Boys at high voltage in the future to any point on the system.
>>
>>MAJOR DISADVANTAGE:  Unfortunately the present US version of the SI 
>>only allows 2 SIs stacked in split phase with no parallel stacking 
>>for bigger systems.  So that means we'd have two sets of the 
>>batteries, two sets of AC, 3-wire outputs at 56 amps each instead 
>>of of one at 120.  And the load peaks would have to always be 
>>distributed amongst the outputs to handle the peaks and overall 
>>loads.  It also means that any expansion faces the same hassle.
>>
>>PROPOSED SYSTEM #2:  8 Stack of Outbacks located near the array.
>>
>>MAJOR DISADVANTAGES:   A 1200' line and ditching would have to be 
>>added from gen site to array/power shed to house to feed the 
>>inverters from the generator with major sized wire as all power 
>>would have to be fed through inverters. - 120 Amps of 240 now 
>>having to go 1200'  (over 4% drop using 750 MCM)  when there are 
>>already existing AC wire runs are from Generator to house.  So all 
>>the power would have to run around two sides of the triangle to get 
>>to the load and the existing 600' of 3/0 Al 3 wire goes unused. . 
>>Also another AC line would have to be run back to feed the loads 
>>that I didn't mention that are on the other side of the generator 
>>from the house. Getting into massive AC wire runs here just because 
>>we are limited to the MX60s voltage for charging from the array and 
>>hence need to keep the  power shed close to the array.  Future wind 
>>and hydro inputs will work if we use a HV/LV scheme (240-600 V 
>>Alternator rectified to 48VDC at power shed) as they are closer to 
>>the array.  All future PV or other input expansion would have to 
>>come to this power shed, however - not be able to tie in anywhere 
>>on the AC "grid" as in the SI/SB system.
>>
>>SO I SEE THIS SITUATION AS THE PERFECT ILLUSTRATION OF THE NEED AND 
>>LIMITIATIONS IN THE INDUSTRY.  I do alot of small village power 
>>systems.  AC coupling is the cat's pajamas,  IF you have the 
>>european SI!
>>
>>On the SI side, the limitation of no parallel stacking is a MAJOR 
>>problem.  Kind of eliminates the whole point of AC coupling for 
>>larger systems - the flexibility to grow systems and feed them 
>>anywhere on the AC side with high voltage.  On the Outback side no 
>>AC coupling means that array and generator have to be close to the 
>>power shed and all future inputs are relegated to DC (which for 
>>wind and hydro is not too bad as you can do them high voltage AC 
>>and then rectify and transformer down).   But in many situations 
>>(not this one) having to always come back to one point instead of 
>>being to tie into the AC grid anywhere on the property will always 
>>be a major limitation.
>>
>>SO QUESTIONS AND POSSIBLE SOLUTIONS:
>>
>>1) I could try to do a SB running into an Outback stack located 
>>near the gen shed.  But we are talking 12 KWs of PV here - with no 
>>software talking - this could get hairy.  And I'd have to run 
>>control wire between SB and Outback for the relay.
>>
>>2)  Any charge controls out there that go higher voltage than the 
>>MX60s 72 volt nominal?  If I run the calcs of trying to run even 
>>the 72 volt (figure 84 volt input?) 600 feet so I can have my power 
>>shed at my genset, we are looking at over 750mcm wire!  not going 
>>there. What we need is a 500 VDC MPPT charge control that feeds 
>>into 48!
>>
>>3)  I can play with locating the power shed at various places 
>>within the triangle . . . . . but all the problems are still there 
>>- just trading off different sets of problems for other sets.
>>
>>4)  According to SMA an updated version of the SI is planned - 
>>hopefully with increased parallel stacking (and hopefully 
>>throttling of inverters like the Outbacks).  According to rumors, 
>>Outback is looking at AC coupling.  In the meantime my client is 
>>burning through THOUSANDS of dollars of diesel running that 30 KW 
>>genset to just to cook breakfast.  The race is on.  In the 
>>meantime, what do I do?!!!!!
>>
>>Well if you read this far, this post should raise some discussion! 
>>Let's be gentle on both SMA and Outback now folks!  It's great we 
>>can even be having this discussion thanks to the great innovation 
>>from both companies!
>>
>>Have at if if you will!
>>
>>Thanks!
>>
>>Jeff Clearwater
>>Village Power Design
>>
>>
>>
>>
>>
>>
>>
>>>Hi Jeff,
>>>
>>>About 6 months ago we set up an off grid system with two SI's coupled with
>>>three SB3800's and about 9 kW of Sanyo HIT's spread evenly amongst the
>>>SB3800's.    It's all set up as 120/240 with no transformers.
>>>
>>>No problems after we upgraded the software.  Everything talks and gets along
>>>fine.
>>>
>>>Travis Creswell
>>>Ozark Energy Services
>>>
>>>-----Original Message-----
>>>From: Jeff Clearwater [mailto:clrwater at earthlink.net]
>>>Sent: Friday, February 03, 2006 4:37 PM
>>>To: RE-wrenches at topica.com
>>>Subject: Sunny Island Experience? [RE-wrenches]
>>>
>>>Hi Folks,
>>>
>>>Anyone running SBs into SIs with AC coupling?  What's your experience been?
>>>
>>>I talked to SMA a few times about running WBs and HBs  into SIs, but
>>>was advised that that would be smoother when SI version 2 is released
>>>- but they didn't know when that would be.  Anyone try that yet?
>>>
>>>Apparently version 2 may have true multimastering as well - where
>>>you can have different SIs on different points on the AC network,
>>>have SBs feeding in and not have the two SIs confuse or compete with
>>>each other.  The present SIs don't support that.
>>>
>>>Thanks in advance!
>>>
>>>Jeff C.
>>>Village Power Design
>>>--
>>>~~~~~~~~~~~~~
>>>Jeff Clearwater
>>>Village Power Design Associates
>>>Sustainable Energy & Water Solutions for Home & Village
>>>http://www.villagepower.com
>>>gosolar at villagepower.com
>>>
>>>530-470-9166
>>>877-SOLARVillage
>>>877-765-2784
>>>425 Nimrod St.
>>>Nevada City, CA 95959
>>>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~`~
>>
>>
>>
>
>Your free subscription is supported by today's sponsor:
>-------------------------------------------------------------------
>Unlimited International Calls to Certain Counties with
>Packet8. Our International VoIP Plans offer subscribers a
>easier way to save on overseas calls. 1-877-868-1135
>http://click.topica.com/caaer6Ibz8Qcsa9hBZif/ TMONE
>-------------------------------------------------------------------
>
>- - - -
>To send a message: RE-wrenches at topica.com
>
>Archive of previous messages: http://lists.topica.com/lists/RE-wrenches/read
>
>List rules & how to change your email address: 
>www.mrsharkey.com/wrenches/etiquette.php
>
>Check out participant bios: www.mrsharkey.com/wrenches/
>
>Hosted by Home Power magazine
>
>Moderator: michael.welch at homepower.com


-- 
~~~~~~~~~~~~~
Jeff Clearwater
Village Power Design Associates
Sustainable Energy & Water Solutions for Home & Village
http://www.villagepower.com
gosolar at villagepower.com

530-470-9166
877-SOLARVillage
877-765-2784
425 Nimrod St.
Nevada City, CA 95959
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~`~

</x-flowed>



More information about the RE-wrenches mailing list